Post on 28-Oct-2014
Globalizing the Cost of
Capital and Capital Budgeting at
AES
A007Pratik DaveA013 Carol loboA015Jay MaliaA025Kartiki Thorave A030AmitMehra
Founded in 1981 (Applied Energy Services) by Roger Sant & Dennis Bakke1983: 1st cogeneration facility is built in Houston, Texas 1988: Net income = $1.6 million 1991: AES goes public, net income = $42.6 million 1991-1992: AES initiates international expansion 1996-1998: estimated 80%-85% capital investment is overseas 2000: Revenue = $6.206 billion
Net Income = $795 million In 2003:
Leading independent supplier of electricity in the world $33 Billion in asset (e.g. Power plants, generation facility, other energy
related businesses) stretched across 30 countries and 5 continents
Revenues by Line of Business and Geographic Region - 2002
Source: AES Corporation, 2002 Annual Report
Typical Investment Structure
What Happened?It's recipe for success (international exposure) became their
recipe for disaster Much of AES' expansion took place in developing countries
(there was more unmet demand vs. developed countries) Main factors:
Devaluation of key South American currencies Argentine, Brazilian, Venezuelan currency crises Adverse changes in energy regulatory requirements Government mandated energy rationing and competition Decline in energy commodity
AES Stock Price History
Market cap in 2002 fell 95% to $1.6 billion@ $1/share
Market cap in2000 reached $28 billion @ $70/share
Capital Budgeting method used historically by AES?
What’s good and bad about it?
Simple Domestic Finance Framework 12% discount rate was used for all contract generation
projects All nonrecourse debt was regarded as goodAll dividend flows from projects were considered
equally risky Project was evaluated by the equity discount rate for
the dividends from the projectFair assumption because businesses had similar capital
structures Most risks could be hedged in the domestic market
Same Model was Exported Overseas
Worked well initially, when they first expanded to Northern Ireland
Had many of the same characteristics as domestic projects
Model became increasingly strained in Brazil and Argentina
Hedging key exposures was not feasible(currency risk , regulatory risk, etc.)
How did AES deal with it? Rob Venerus, director of Corporate Analysis &
Planning questioned whether the traditional CAPM would suffice
AES owned businesses in poorly integrated capital markets
WACC was not working due to the events that were occurring internationally
He did not advocate the use of a world CAPM He did not advocate the use a local CAPM either
Countries such as Tanzania and Georgia did not have any meaningful capital markets
New ModelStep 1
Calculate the cost of equity using U.S. market data for each of AES' projects
Average the unlevered equity betas from comparable U.S. companies
Relever the beta to reflect the capital structure of each of AES' projects
Cost of equity = Rf + (Rm – Rf) β
New ModelStep 2
Calculate the cost of debt by adding the U.S. risk free rate and a "default spread"
Cost of Debt = Rf + Default Spread The "default spread" is based on the
relationship between EBIT ratios for comparable companies and their cost of debt.
New ModelStep 3
Add the sovereign spread to both the cost of equity and the cost of debt
this accounts for country-specific market risk, which is the difference between local government bond yields and corresponding U.S. Treasury yields.
These steps allow AES to calculate a WACC that reflects the systematic risk associated with each project in its local market.
New ModelMost of these local markets are developing markets where
"access to capital was limited and information less than perfect" hence project specific risk could not be diversified away .
Example of project-specific risk: There are 2 hydro plants in Brazil that are identical in every aspect, except for the rivers that feed them. River #1 produces cash flows that vary +/- 50%, River #2 produces cash flows by +/- 10%.
If they are financed by 100% equity, CAPM says they are worth the same.
Rob Venerus thought this was unconvincing
"Project-specific risk" must be accounted for!
Seven types of "Project-specific risk"
1. Operational/Technical 2. Counterparty credit/performance 3. Regulatory 4. Construction 5. Commodity 6. Currency 7. Contractual Enforcement/Legal
Weights estimated from AES' ability to anticipate and mitigate risk. Then given a grade between 0 (lowest exposure) and 3 (highest exposure), multiplied by their weights to yield a "business-specific risk score"
WACC Calculation
Used to calculate an adjustment to the initial cost of capital
0 = no adjustment to WACC 1 = +500 basis points (5%) 2 = +1000 basis points (10%) 3 = +1500 basis points (15%)
Add a business-specific risk to WACC & the final adjusted to WACC is arrived.
Lets understand the New Model by taking Project of 2 country as under:
Lal Pir Project – Pakistan
Red Oak Project - USA
Categories of Risk Weight
Grade for Lal
Pir
Risk Scores (grade x weight)
Grade for Red Oak
Risk Scores
(grade x weight)
Operational / Technical 3.50% 1 0.035 2 0.07
Counterparty Credit / Performance 7.00% 1 0.07 3 0.21
Regulatory 10.50% 2 0.21 0 0
Construction 14.50% 0 0 0 0
Commodity 18.00% 1 0.18 2 0.36
Currency 21.50% 2 0.43 0 0Contractual Enforcement / Legal 25.00% 2 0.5 0 0Sum of individual scores = business specific risk score 1.425 0.64
Premium (in bp) 712.5 320
Premium (in bp) in % 7.13 3.2
Risk Score Calculation for Lal Pir Project & Red Oak
Discount Rate Adjustment: USA v/s Pakistan
Discount rate is adjusted based on the total risk score of the country
In Pakistan currency, regulatory and legal risks are significantly high
Operational, counterparty and commodity risks are higher in USA as compared to PakistanAdjusted WACC for Pakistan (23.08%) is much higher as opposed to that of USA (9.66%)Pakistan is inherently a riskier country to invest in as opposed to the USA and any investments made in this region would have to cross a higher hurdle rate than if they were made in the US region.
Business / Project Country Line of Business Adjusted WACCRed Oak USA CG 9.66%Ottana Italy CS 10.77%Gener Chile CG 12.63%Kelvin South Africa CG 15.18%Drax United Kingdom CS 16.35%
Haripur Bangladesh CG 16.88%OPGC India CG 18.11%
Rivnoblenergo Ukraine GD 18.58%Lal Pir Pakistan CG 23.08%
Uruguaiana Brazil CG 25.15%Eletropaulo Brazil LU 25.26%
Los Mina Dominican Republic CG 27.44%Telasi Georgia GD 28.51%
Andres Dominican Republic CG 29.94%Caracoles Argentina CS 31.36%
Range of discount rates that AES can use around world
Suggestion & Recommendation
New model to value cost and risk should be implemented by AES.
It gives the company a more realistic projection of the risks that they may face with projects that they take on internationally.
Risks such as political, economic, country-specific and business-specific risks are now considered, where in the previous model they were neglected.